This is a follow-up to some related questions: 46770 and 60600.
I'd like to make a DistributionChart where the elements are histograms with my own custom settings. For example, I'd like to be able to take:
data = WeightedData[Range[10], #] & /@ RandomInteger[20, {5, 10}];
and make a DistributionChart where each element looks like, say,
Histogram[data[[i]],10,{"Log","Probability"}][[1]]
for the appropriate value of the iterator, i (ideally with the histogram bars centered rather than edge-aligned). I tried something similar to this answer:
Module[
{cef :=
Histogram[#2,10,{"Log", "Probability"}, BarOrigin -> Left][[1]] /.
RectangleBox[{x0_, y0_}, {x1_, y1_}, z___] :>
RectangleBox[
{(- x1 + #[[1, 1]] + #[[1,2]])/2, y0},
{(x1 + #[[1, 1]] + #[[1, 2]])/2, y1}, z] &},
DistributionChart[data, ChartElementFunction -> cef]]
The code above almost works, but because the bars of the histograms can have length greater than 1, they can overlap. (It works fine if {"Log", "Probability"} is changed to "Probability".) I can fix this on a case-by-case basis by scaling down the bar lengths by some factor, but is there a way to do it automatically?
